tape or what?

[Carmel Bruno]

Now that Red and other cameras has been out for donkeys years and tape is dead, where is everybody archiving their data? I have hours and hours of tapes that I still get snippets from when needed. Recently I have started editing for a local TV station who use drives and already I have terabytes of their data that I am working with. No problem so far, because once I master, I can ditch their data. But my footage is another matter. If I do the big move to a no tape environment, what am I going to do to store the lots of hours of footage? Archiving stuff that I might need in the next ten years or so?

 

[Dan Hudgins]

Film

Film

Now that Red and other cameras has been out for donkeys years and tape is dead, where is everybody archiving their data? I have hours and hours of tapes that I still get snippets from when needed. Recently I have started editing for a local TV station who use drives and already I have terabytes of their data that I am working with. No problem so far, because once I master, I can ditch their data. But my footage is another matter. If I do the big move to a no tape environment, what am I going to do to store the lots of hours of footage? Archiving stuff that I might need in the next ten years or so?

Celco says use 35mm black and white movie film,

http://www.celco.com/cinesafe4k.asp

 

[Carmel Bruno]

Celco says use 35mm black and white movie film,

http://www.celco.com/cinesafe4k.asp

but that is only for 500 years...
what if I need a particular piece if footage in 501 years time?

 

[Erik Franzén]

but that is only for 500 years...
what if I need a particular piece if footage in 501 years time?

[...] black and white film will last in excess of 500 years

I think you're more than safe ;-)

 

[Rob Ruffo]

LTO = Good, now cheap solution

 

[Andrew M.]

I like the idea of using film as an optical medium of storing digital data in form of zeros and ones.

Kind of barcode exposed on the film.

Imagine all these film manufacturing plants going idle soon will have some use in the film industry and not only film industry for storing digital data.

The problem though is the speed of access. It is linear media.

 

[Andy Jarosz]

I like the idea of using film as an optical medium of storing digital data in form of zeros and ones.

Kind of barcode exposed on the film.

Imagine all these film manufacturing plants going idle soon will have some use in the film industry and not only film industry for storing digital data.

The problem though is the speed of access. It is linear media.

A post of mine from a while back...

http://reduser.net/forum/showpost.php?p=656944&postcount=9

 

[Michael Olsen]

My understanding is that tape may be "dead" as an acquisition format, but that the LTO tapes offer a great bang-for-buck backup solution.

And, of course, optically storing it as film is always a valid option.

 

[Andrew M.]

My understanding is that tape may be "dead" as an acquisition format, but that the LTO tapes offer a great bang-for-buck backup solution.

And, of course, optically storing it as film is always a valid option.

Looks like some people are already looking to store very high densities onto the film.

http://iopscience.iop.org/0957-4484/13/6/307;jsessionid=FB1DAB58CA08B7445C5CABCB88552FE9.c3

I guess with a little research investment Kodak may rebound making film that can be used to archive digital movies. What an irony, but it may work.

Archives do not need frequent random access. Linear storage would be OK for it.


Someone can even modify existing projectors to read such digital storrage movies in the theatres and send such retrieved data straight into 4K projectors.




Just kidding.....

 

[Gavin Greenwalt]

I prefer redundancy to reliability myself. Look at the cockroach. Would you rather be a million cockroaches or 1 clever rat? Sure the rat is a few thousand times more intelligent... but good luck killing all the cockroaches.

Every format is prone to failure. Fire. Water. Mishandling. Apathy. All are dangerous to data. The best long term storage medium is duplication. If you make a million of anything you're bound to have one survive.

Digital isn't *that* unreliable. With a little effort you can far more cheaply store data on a hdd just as reliably for less than optical archiving. And unlike optical you have Moore's law on your side.

If 1TB HDDs cost $50 today in 3 years they'll cost $10. In 6 years 1TB of storage will cost $2.50. In 6 years it'll still cost several thousand dollars to archive a film optically. Sure it'll last "500 years". But would you rather have 2 copies rated to survive 500 years in a perfect hypothetical vault for $10k or 4,000 copies for the same price spread across the world? If after 200 years 95% of them failed 200 would still work and it only takes one of them to move to the next best storage medium. In 100 years I can guarantee you we'll have storage that will hold the data for millenia without losing a bit.

You don't have to store it for a million years. You just need to store it until we develop a millenial archiving format. Which assumes somewhat egotistically and irrationally that someone will care about our data in 50 years let alone 500.

EDIT: By the way I still have HDDs from the 90s which function perfectly and a stack of CDs which I backed up my HDD at the time to. About 10% of them no longer work a decade later (almost entirely from scratches and careless handling). I can't find many of my negatives from the same time period. In this last week my Dad lost almost all of his slides and negatives to a fire. All he has now is the scans. Redundancy triumphs reliability.

 

[Michael Olsen]

I prefer redundancy to reliability myself.

This is an important point, too.

If one is concerned about their archive remaining usable for 500+ years, I have a feeling they would undoubtedly want to have multiple copies of the backup distributed to different physical locations...and at least two very safe places.

There have been a lot of natural disasters, wars, and plain old human errors since the 1500s.

--

This raises a genuine question on my behalf. I have always read that HDD have a relatively short data life-span unless they are occasionally spun up for a few hours. This is usually attributed to the gradual loss of magnetic signature on the platter, and is supposed to reach a catastrophic level in about 1 years time, at which point the data may become useless.

Does anyone here have some real world experience with that? My own do not bear that out - I have had systems drives boot up after several years of being shelved, no questions asked. But that's a pretty small sample.

And then, of course, the bigger question. If I take a 1TB drive and leave it locked in a vault somewhere, what is the chance it will still work in 50, 100, 200, etc? (Not, of course, accounting for hardware advancements which will undoubtedly render "old" formats hard or impossible to access.)

 

[Gavin Greenwalt]

Does anyone here have some real world experience with that? My own do not bear that out - I have had systems drives boot up after several years of being shelved, no questions asked. But that's a pretty small sample.

And then, of course, the bigger question. If I take a 1TB drive and leave it locked in a vault somewhere, what is the chance it will still work in 50, 100, 200, etc? (Not, of course, accounting for hardware advancements which will undoubtedly render "old" formats hard or impossible to access.)

It's my understanding that the problem with idle drives isn't the magnetic signature (which doesn't get rewritten) but the bearings going out.

Which then raises the old cost benefit question again. What's the cost of having a forensic data specialist remount your drive to a good backup? Just because a computer can't see the data doesn't mean you couldn't have a specialist put the spindles on a more sensitive reader (for a price).

in my narrow experience I've never had an old drive that worked when disconnected stop working. And I've popped in some really old drives before.